From marka@isc.org Sat Nov 19 00:46:46 2016 In-Reply-To: Your message of "Fri, 18 Nov 2016 11:17:16 -0000." X-Spam-Status: No, score=-4.3 required=5.0 tests=ALL_TRUSTED,BAYES_00, RP_MATCHES_RCVD autolearn=ham autolearn_force=no version=3.4.0 X-RT-Interface: API References: content-type: text/plain; charset="utf-8" Message-ID: <20161119004633.EB4A25A77B0E@rock.dv.isc.org> X-RT-Original-Encoding: utf-8 Received: from mx.ams1.isc.org (mx.ams1.isc.org [199.6.1.65]) by bugs.isc.org (Postfix) with ESMTP id C9AA071B5A8 for ; Sat, 19 Nov 2016 00:46:45 +0000 (UTC) Received: from zmx1.isc.org (zmx1.isc.org [149.20.0.20]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx.ams1.isc.org (Postfix) with ESMTPS id 3F7771FCAC3 for ; Sat, 19 Nov 2016 00:46:39 +0000 (UTC) Received: from zmx1.isc.org (localhost [127.0.0.1]) by zmx1.isc.org (Postfix) with ESMTPS id 1FFFE16004F for ; Sat, 19 Nov 2016 00:46:38 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by zmx1.isc.org (Postfix) with ESMTP id F02EB160073 for ; Sat, 19 Nov 2016 00:46:37 +0000 (UTC) Received: from zmx1.isc.org ([127.0.0.1]) by localhost (zmx1.isc.org [127.0.0.1]) (amavisd-new, port 10026) with ESMTP id z9VeidsvlYia for ; Sat, 19 Nov 2016 00:46:37 +0000 (UTC) Received: from rock.dv.isc.org (c27-253-115-14.carlnfd2.nsw.optusnet.com.au [27.253.115.14]) by zmx1.isc.org (Postfix) with ESMTPSA id 9688716004F for ; Sat, 19 Nov 2016 00:46:37 +0000 (UTC) Received: from rock.dv.isc.org (localhost [IPv6:::1]) by rock.dv.isc.org (Postfix) with ESMTP id EB4A25A77B0E for ; Sat, 19 Nov 2016 11:46:33 +1100 (EST) Delivered-To: bind9-bugs@bugs.isc.org Subject: Re: [ISC-Bugs #43683] TCP segment handling issue? Return-Path: X-Original-To: bind9-bugs@bugs.isc.org Date: Sat, 19 Nov 2016 11:46:33 +1100 X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on mx.ams1.isc.org To: bind9-bugs@isc.org From: "Mark Andrews" RT-Message-ID: Content-Length: 994 In message , "Ray Bellis via RT" writes: > Via kato@wide.ac.jp: > > bind-9.11.0-p1 seems to me that it applys TCP_MAXSEG to 1220byte even > in IPv4, resulting TCP transmission a little bit less efficient. At > around line 3698 of lib/isc/unix/socket.c, the code should be executed > only if protocol family is AF_INET6: > > if (NEWCONNSOCK(dev)->pf == AF_INET6) { > use_min_mtu(NEWCONNSOCK(dev)); > set_tcp_maxseg(NEWCONNSOCK(dev), 1280 - 20 - 40); > } > > > This relates to the survey done by Geoff Huston and reported on his blog > earlier this week. And there is nothing wrong with this. We really don't want PMTUD to happen over IPv4 or IPv6. Setting the max seg to 1220 avoids triggering PMTUD for IPv6 over IPv4 and IPv4 over IPv6. If anything the pf test should be removed. Mark -- Mark Andrews, ISC 1 Seymour St., Dundas Valley, NSW 2117, Australia PHONE: +61 2 9871 4742 INTERNET: marka@isc.org